DOCSLIB.ORG

Juliafem — Open Source Solver for Both Industrial and Academia Usage

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Juliafem — Open Source Solver for Both Industrial and Academia Usage Rakenteiden Mekaniikka (Journal of Structural Mechanics) Vol. 50, No 3, 2017, pp. 229 { 233 https://rakenteidenmekaniikka.journal.fi/index https://doi.org/10.23998/rm.64224 c The author(s) 2017. Open access under CC BY-SA 4.0 license. JuliaFEM | open source solver for both industrial and academia usage Tero Frondelius1 and Jukka Aho Summary. The JuliaFEM software library is a framework that allows for the distributed processing of large Finite Element Models across clusters of computers using simple programming models. It is designed to scale up from single servers to thousands of machines, each offering local computation and storage. The basic design principle is: Everything is non-linear. All physics models are non-linear from which the linearizations are made as special cases. This is a work in progress. Thus, if you share the vision, contribute and join the community. Key words: JuliaFEM, FEM, open source, Julia, julialang, natural frequency, www.juliafem.org Received 1 June 2017. Accepted 11 August 2017. Published online 21 August 2017. Introduction The JuliaFEM project was started in May 2015 and is MIT licensed [6]. However, the story actually starts from years back of studying and using other open-source FEM packages. The findings were mainly divided into two categories: namely first, academic projects with pure academic goals and parallelism missing and, second, fast parallel codes that were so hard to start using and typically written-in statically compiled languages. Customizing software written using such languages requires significant developer time, thus sacrificing human convenience and even more importantly, typically very limited developer resources for software execution speed. JuliaFEM is an open-source finite element solver written in the Julia programming language, which is comprehensibly described in [3]. All the same reasoning are true for why Julia and why JuliaFEM, in other words why new programming language was needed. Reference [3] answers the first question throughly, thus the authors encourage readers to read it or visiting https://julialang.org/. JuliaFEM enables flexible simulation models, takes advantage of the scripting language interface which is easy to learn and embrace. In addition, it is a real programming environment where simulation can be combined with other analyses and work-flows. These features introduce a place for testing new ideas and simulation models to the academic world. 1Corresponding author. [email protected] 229 Julia introduces the following features: JIT/LLVM [10], multiple dispatch, metapro- gramming, very powerful macros (full programming language with Julia syntax), static and dynamic typing. The speed of Julia is really close to the speed of C/FORTRAN code. Julia shows that one can have machine performance without sacrificing human convenience. Julia combines expertise from the diverse fields of computer science and computational science to create a new approach to numerical computing. Julia is de- signed to be easy and fast to use. One of the key features of Julia is multiple dispatch, a technique from computer science which automatically picks the right algorithm for the right circumstance [3]. Julia provides a wide range of libraries and a very skilled community in both skills and education, which enables a high level of abstraction in the programming code, such as using the automatic differentiation library ForwardDiff [16] for calculating gradients, jacobians and hessians for all types of field data. Using automatic differentiation in a FEM package is not a new idea; for example FEniCS [1], an open-source FEM solver, uses it comprehensively. Another important feature of the open-source software development is the method of building on top of existing open-source work. Thus JuliaFEM offers an io- interface to CODE ASTER [17], also an open source FEM solver. Developing JuliaFEM also encourages good practices, starting from unit testing both for smaller and larger functions and continuing to full integration testing of different platforms. At the moment, users can perform the following analyses with JuliaFEM: elasticity, thermal, eigenvalue, contact mechanics, and quasi-static solutions. For visualization, JuliaFEM uses Paraview [2] which prefers XDMF [13] file format using XML to store light data and HDF [18] to store large data-sets, which is more or less the open-source standard. Vision On one hand, the vision of the JuliaFEM includes the opportunity for massive paralleliza- tion using multiple computers with MPI and threading as well as cloud computing re- sources in Amazon, Azure [7] and Google cloud services together with a company internal server [9]. And on the other hand, the real application complexity including the simula- tion model complexity as well as geometric complexity, see some examples in [4,8, 19]. Not to forget that the reuse of the existing material models [11] as well as the whole simulation models are considered crucial features of the JuliaFEM package [5, 12]. Recreating the wheel again is definitely not anybody's goal, and thus we try to use and embrace good practices and formats as much as possible. We have implemented Abaqus/Calculix [21] input-file format support and maybe will in the future extend to Other FEM solver formats. Using modern development environments encourages the user towards fast development time and high productivity. For developing and creating new ideas and tutorials, we have used Jupyter notebooks [15] to make easy-to-use handouts. The user interface for JuliaFEM is Jupyter Notebook [15], and Julia language itself is a real programming language. This makes it possible to use JuliaFEM as a part of a bigger solution cycle, including for example data mining, automatic geometry modifications, mesh generation, solution, and post-processing and enabling efficient optimization loops. In the next section a concrete example including statistical inference combined to FEM calculation is shown. 230 Big numerical example Typical examples in industrial applications include non-linear solid mechanics [11], contact mechanics [20], finite strains [14], and fluid structure interaction problems. Here there is some simulation of machine parts having different amounts of elements and DOF's for comparison reasons. These simulations take a lot of computational resources, and here are specs of the used hardware: 24 x Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz with 512 GB total memory. Table 1. This table demonstrates the solution times for some real industrial size FEM models DOFS Assembly time (s) Solution time (s) Total time (human readable) 3.0 M 458 312 16 min 10.8 M 3257 3255 2 h 4 min 12.6 M 3654 6318 3 h 4 min Figure 1. Real industrial size FEM model. References [1] Martin S. Alnæs, Jan Blechta, Johan Hake, August Johansson, Benjamin Kehlet, Anders Logg, Chris Richardson, Johannes Ring, Marie E. Rognes, and Garth N. Wells. The FEniCS project version 1.5. Archive of Numerical Software, 3(100), 2015. URL https://doi.org/10.11588/ans.2015.100.20553. [2] Utkarsh Ayachit et al. The ParaView guide: A parallel visualization application, kitware, 2015. [3] Jeff Bezanson, Alan Edelman, Stefan Karpinski, and Viral B. Shah. Julia: A fresh approach to numerical computing. SIAM Review, 59(1):65{98, 2017. URL https: //doi.org/10.1137/141000671. [4] Tero Frondelius, Pasi Halla-aho, and Antti M¨antyl¨a.Crankshaft development with virtual engine modelling. In CIMAC Congress Helsinki, 2016. 231 [5] Jussi G¨o¨os, Anton Lepp¨anen, Antti M¨antyl¨a, and Tero Frondelius. Large bore connecting rod simulations. Rakenteiden Mekaniikka, 50(3):275{278, 2017. URL https://doi.org/10.23998/rm.64658. [6] Open Source Initiative et al. The MIT license, 2006. [7] Roger Jennings. Cloud Computing with the Windows Azure Platform. Wrox Press Ltd., Birmingham, UK, UK, 2009. ISBN 0470506385, 9780470506387. [8] Juho K¨onn¨o,Tero Frondelius, Thomas Resch, and Maria Jose Santos-Descalzo. Sim- ulation based grid compliance. In CIMAC Congress Helsinki, 2016. [9] Juho K¨onn¨o,Hannu Tienhaara, and Tero Frondelius. W¨artsil¨adigital design plat- form. Rakenteiden Mekaniikka, 50(3):234{238, 2017. URL https://doi.org/10. 23998/rm.64621. [10] Chris Lattner and Vikram Adve. LLVM: A compilation framework for lifelong pro- gram analysis and transformation. pages 75{88, San Jose, CA, USA, Mar 2004. [11] Anton Lepp¨anen,Asko Kumpula, Joona Vaara, Massimo Cattarinussi, Juho K¨onn¨o, and Tero Frondelius. Thermomechanical fatigue analysis of cylinder head. Rakentei- den Mekaniikka, 50(3):182{185, 2017. URL https://doi.org/10.23998/rm.64743. [12] Antti M¨antyl¨a,Jussi G¨o¨os,Anton Lepp¨anen,and Tero Frondelius. Large bore engine connecting rod fretting analysis. Rakenteiden Mekaniikka, 50(3):239{243, 2017. URL https://doi.org/10.23998/rm.64914. [13] E Mark et al. Enhancements to the extensible data model and format (xdmf). In DoD High Performance Computing Modernization Program Users Group Conference, 2007, pages 322{327. IEEE, 2007. [14] Mikael Nyberg, Antti M¨antyl¨a,and Tero Frondelius. Explosion simulation of pres- surized components. Rakenteiden Mekaniikka, 50(3):198{200, 2017. URL https: //doi.org/10.23998/rm.65076. [15] Fernando P´erezand Brian E. Granger. IPython: a system for interactive scientific computing. Computing in Science and Engineering, 9(3):21{29, May 2007. ISSN 1521-9615. URL https://doi.org/10.1109/MCSE.2007.53. [16] J. Revels, M. Lubin, and T. Papamarkou. Forward-mode automatic differentiation in Julia. arXiv:1607.07892 [cs.MS], 2016. URL https://arxiv.org/abs/1607.07892. [17] J. Russell and R. Cohn. Code Aster. Book on Demand, 2012. ISBN 9785511435350. URL https://books.google.fi/books?id=9iN8MAEACAAJ. [18] The HDF Group. Hierarchical data format version 5, 2000-2010. URL http://www. hdfgroup.org/HDF5. [19] Ilkka V¨ais¨anen,Antti M¨antyl¨a,Antti Korpela, Teemu Kuivaniemi, and Tero Fron- delius. Medium speed engine crankshaft analysis. Rakenteiden Mekaniikka, 50(3): 341{344, 2017.
Load more

Recommended publications
  • Parallel Data Analysis Directly on Scientific File Formats
    Parallel Data Analysis Directly on Scientific File Formats Spyros Blanas 1 Kesheng Wu 2 Surendra Byna 2 Bin Dong 2 Arie Shoshani 2 1 The Ohio State University 2 Lawrence Berkeley National Laboratory [email protected] {kwu, sbyna, dbin, ashosani}@lbl.gov ABSTRACT and physics, produce massive amounts of data. The size of Scientific experiments and large-scale simulations produce these datasets typically ranges from hundreds of gigabytes massive amounts of data. Many of these scientific datasets to tens of petabytes. For example, the Intergovernmen- are arrays, and are stored in file formats such as HDF5 tal Panel on Climate Change (IPCC) multi-model CMIP-5 and NetCDF. Although scientific data management systems, archive, which is used for the AR-5 report [22], contains over petabytes such as SciDB, are designed to manipulate arrays, there are 10 of climate model data. Scientific experiments, challenges in integrating these systems into existing analy- such as the LHC experiment routinely store many gigabytes sis workflows. Major barriers include the expensive task of of data per second for future analysis. As the resolution of preparing and loading data before querying, and convert- scientific data is increasing rapidly due to novel measure- ing the final results to a format that is understood by the ment techniques for experimental data and computational existing post-processing and visualization tools. As a con- advances for simulation data, the data volume is expected sequence, integrating a data management system into an to grow even further in the near future. existing scientific data analysis workflow is time-consuming Scientific data are often stored in data formats that sup- and requires extensive user involvement.
    [Show full text]
  • A Metadata Based Approach for Supporting Subsetting Queries Over Parallel HDF5 Datasets
    A Metadata Based Approach For Supporting Subsetting Queries Over Parallel HDF5 Datasets Thesis Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University By Vignesh Santhanagopalan, B.S. Graduate Program in Computer Science and Engineering The Ohio State University 2011 Thesis Committee: Dr. Gagan Agrawal, Advisor Dr. Radu Teodorescu ABSTRACT A key challenge in scientific data management is to manage the data as the data sizes are growing at a very rapid speed. Scientific datasets are typically stored using low-level formats which store the data as binary. This makes specification of the data and processing very hard. Also, as the volume of data is huge, parallel configurations must be used to process the data to enable efficient access. We have developed a data virtualization approach for supporting subsetting queries on scientific datasets stored in native format. The data is stored in Hierarchical Data Format (HDF5) which is one of the popular formats for storing scientific data. Our system supports SQL queries using the Select, From and Where clauses. We support queries based on the dimensions of the dataset and also queries which are based on the dimensions and attributes (which provide extra information about the dataset) of the dataset. In order to support the different types of queries, we have the pre-processing and post-processing modules. We also parallelize the selection queries involving the dimensions and the attributes. Our system offers the following advantages. We provide SQL like abstraction for specifying subsets of interest which is a powerful mechanism.
    [Show full text]
  • HUDDL for Description and Archive of Hydrographic Binary Data
    University of New Hampshire University of New Hampshire Scholars' Repository Center for Coastal and Ocean Mapping Center for Coastal and Ocean Mapping 4-2014 HUDDL for description and archive of hydrographic binary data Giuseppe Masetti University of New Hampshire, Durham, [email protected] Brian R. Calder University of New Hampshire, Durham, [email protected] Follow this and additional works at: https://scholars.unh.edu/ccom Recommended Citation G. Masetti and Calder, Brian R., “Huddl for description and archive of hydrographic binary data”, Canadian Hydrographic Conference 2014. St. John's, NL, Canada, 2014. This Conference Proceeding is brought to you for free and open access by the Center for Coastal and Ocean Mapping at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Center for Coastal and Ocean Mapping by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. Canadian Hydrographic Conference April 14-17, 2014 St. John's N&L HUDDL for description and archive of hydrographic binary data Giuseppe Masetti and Brian Calder Center for Coastal and Ocean Mapping & Joint Hydrographic Center – University of New Hampshire (USA) Abstract Many of the attempts to introduce a universal hydrographic binary data format have failed or have been only partially successful. In essence, this is because such formats either have to simplify the data to such an extent that they only support the lowest common subset of all the formats covered, or they attempt to be a superset of all formats and quickly become cumbersome. Neither choice works well in practice.
    [Show full text]
  • Towards Interactive, Reproducible Analytics at Scale on HPC Systems
    Towards Interactive, Reproducible Analytics at Scale on HPC Systems Shreyas Cholia Lindsey Heagy Matthew Henderson Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Drew Paine Jon Hays Ludovico Bianchi Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Devarshi Ghoshal Fernando Perez´ Lavanya Ramakrishnan Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Abstract—The growth in scientific data volumes has resulted analytics at scale. Our work addresses these gaps. There are in a need to scale up processing and analysis pipelines using High three key components driving our work: Performance Computing (HPC) systems. These workflows need interactive, reproducible analytics at scale. The Jupyter platform • Reproducible Analytics: Reproducibility is a key com- provides core capabilities for interactivity but was not designed ponent of the scientific workflow. We wish to to capture for HPC systems. In this paper, we outline our efforts that the entire software stack that goes into a scientific analy- bring together core technologies based on the Jupyter Platform to create interactive, reproducible analytics at scale on HPC sis, along with the code for the analysis itself, so that this systems. Our work is grounded in a real world science use case can then be re-run anywhere. In particular it is important - applying geophysical simulations and inversions for imaging to be able reproduce workflows on HPC systems, against the subsurface.
    [Show full text]
  • Parsing Hierarchical Data Format (HDF) Files Karl Nyberg Grebyn Corporation P
    Parsing Hierarchical Data Format (HDF) Files Karl Nyberg Grebyn Corporation P. O. Box 47 Sterling, VA 20167-0047 703-406-4161 [email protected] ABSTRACT This paper presents a description of the creation of a library to parse Hierarchical Data Format (HDF) Files in Ada. It describes a “work in progress” with discussion of current performance, limitations and future plans. Categories and Subject Descriptors D.2.2 [Design Tools and Techniques]: Software Libraries; E.2 [Data Storage Representations]: Object Representation; I.2.10 [Vision and Scene Understanding]: Representations, data structures, and transforms. General Terms Algorithms, Performance, Experimentation, Data Formats. Keywords Ada, HDF. 1. INTRODUCTION Large quantities of scientific data are published each year by NASA. These data are often accompanied by metadata files that describe the contents of individual files of the data. One example of this data is the ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) [1]. Each file of data (consisting of satellite imagery) in HDF (Hierarchical Data Format) [2] is accompanied by a metadata file in XML (Extensible Markup Language) [3], encoded according to a published DTD (Document Type Description) that indicates the components and types of data in the metadata file. Each ASTER data file consists of an image of the satellite as it passes over the earth [4]. Information on the location of the data collected as the satellite passes is contained in the metadata file. Over time, multiple images of the same location on earth are obtained. For many purposes of analysis (erosion, building patterns, deforestation, glacier movement, etc.), these images of the same location are compared over time.
    [Show full text]
  • A New Data Model, Programming Interface, and Format Using HDF5
    NetCDF-4: A New Data Model, Programming Interface, and Format Using HDF5 Russ Rew, Ed Hartnett, John Caron UCAR Unidata Program Center Mike Folk, Robert McGrath, Quincey Kozial NCSA and The HDF Group, Inc. Final Project Review, August 9, 2005 THG, Inc. 1 Motivation: Why is this area of work important? While the commercial world has standardized on the relational data model and SQL, no single standard or tool has critical mass in the scientific community. There are many parallel and competing efforts to build these tool suites – at least one per discipline. Data interchange outside each group is problematic. In the next decade, as data interchange among scientific disciplines becomes increasingly important, a common HDF-like format and package for all the sciences will likely emerge. Jim Gray, Distinguished “Scientific Data Management in the Coming Decade,” Jim Gray, David Engineer at T. Liu, Maria A. Nieto-Santisteban, Alexander S. Szalay, Gerd Heber, Microsoft, David DeWitt, Cyberinfrastructure Technology Watch Quarterly, 1998 Turing Award Volume 1, Number 2, February 2005 winner 2 Preservation of scientific data … the ephemeral nature of both data formats and storage media threatens our very ability to maintain scientific, legal, and cultural continuity, not on the scale of centuries, but considering the unrelenting pace of technological change, from one decade to the next. … And that's true not just for the obvious items like images, documents, and audio files, but also for scientific images, … and MacKenzie Smith, simulations. In the scientific research community, Associate Director standards are emerging here and there—HDF for Technology at (Hierarchical Data Format), NetCDF (network the MIT Libraries, Common Data Form), FITS (Flexible Image Project director at Transport System)—but much work remains to be MIT for DSpace, a groundbreaking done to define a common cyberinfrastructure.
    [Show full text]
  • A Very Useful Enhancement for MSC Nastran and Patran
    Tips & Tricks HDF5: A Very Useful Enhancement for MSC Nastran and Patran et’s face it, as FEA engineers we’ve all been As a premier FEA solver that is widely used by the in that situation where the finite element Aerospace and Automotive industries, MSC Nastran L model has been prepared, boundary now takes advantage of the HDF5 file to manage conditions have been checked and the model has and store your FEA data. This can simplify your post been solved, but then it comes time to investigate processing tasks and make data management much the results. It’s not unusual on a large project to easier and simpler (see below). produce different outputs for different purposes: an XDB or OP2 for Patran, a Punch file to pass to So what is HDF5? It is an open source file format, Excel, and the f06 to read the printed output. In data model, and library developed by the HDF Group. addition, your company may have programs that read one of these files to perform special purpose HDF5 has three significant advantages compared to calculations. Managing all these files becomes a previous result file formats: 1) The HDF5 file is project of its own. smaller than XDB and OP2, 2) accessing results is significantly faster with HDF5, and 3) the input and Well let me tell you, there is a better way to do this!! output datablocks are stored in a single, high 82 | Engineering Reality Magazine Before HDF5 After HDF5 precision file. With input and output in the same file, party extensions are available for Python, .Net, and you eliminate confusion and potential errors keeping many other languages.
    [Show full text]
  • Importing and Exporting Data
    Chapter II-9 II-9Importing and Exporting Data Importing Data................................................................................................................................................ 117 Load Waves Submenu ............................................................................................................................ 119 Line Terminators...................................................................................................................................... 120 LoadWave Text Encodings..................................................................................................................... 120 Loading Delimited Text Files ........................................................................................................................ 120 Determining Column Formats............................................................................................................... 120 Date/Time Formats .................................................................................................................................. 121 Custom Date Formats ...................................................................................................................... 122 Column Labels ......................................................................................................................................... 122 Examples of Delimited Text ................................................................................................................... 123 The Load
    [Show full text]
  • Achieving High Performance I/O with HDF5
    Achieving High Performance I/O with HDF5 HDF5 Tutorial @ ECP Annual Meeting 2020 M. Scot Breitenfeld, Elena Pourmal The HDF Group Suren Byna, Quincey Koziol Lawrence Berkeley National Laboratory Houston, TX Feb 6th, 2020 exascaleproject.org Outline, Announcements and Resources INTRODUCTION 2 Tutorial Outline • Foundations of HDF5 • Parallel I/O with HDF5 • ECP HDF5 features and applications • Logistics – Live Google doc for instant questions and comments: https://tinyurl.com/uoxkwaq 3 https://tinyurl.com/uoxkwaq Announcements • HDF5 User Group meeting in June 2020 • The HDF Group Webinars https://www.hdfgroup.org/category/webinar/ – Introduction to HDF5 – HDF5 Advanced Features – HDF5 VOL connectors 4 https://tinyurl.com/uoxkwaq Resources • HDF5 home page: http://hdfgroup.org/HDF5/ • Latest releases: ØHDF5 1.10.6 https://portal.hdfgroup.org/display/support/Downloads ØHDF5 1.12.0 https://gamma.hdfgroup.org/ftp/pub/outgoing/hdf5_1.12/ • HDF5 repo: https://bitbucket.hdfgroup.org/projects/HDFFV/repos/hdf5/ • HDF5 Jira https://jira.hdfgroup.org • Documentation https://portal.hdfgroup.org/display/HDF5/HDF5 5 https://tinyurl.com/uoxkwaq FOUNDATIONS OF HDF5 Elena Pourmal 6 Foundations of HDF5 • Introduction to – HDF5 data model, software and architecture – HDF5 programming model • Overview of general best practices 7 https://tinyurl.com/uoxkwaq Why HDF5? • Have you ever asked yourself: – How do I organize and share my data? – How can I use visualization and other tools with my data? – What will happen to my data if I need to move my application
    [Show full text]
  • Efficient Lidar Point Cloud Data Encoding for Scalable Data Management Within the Hadoop Eco-System
    Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco-system Vo, A. V., Hewage, C. N. L., Russo, G., Chauhan, N., Laefer, D. F., Bertolotto, M., Le-Khac, N-A., & Ofterdinger, U. (2020). Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco- system. In IEEE BigData 2019 Los Angeles, CA, USA (pp. 5644-5653). IEEE . https://doi.org/10.1109/BigData47090.2019.9006044 Published in: IEEE BigData 2019 Los Angeles, CA, USA Document Version: Peer reviewed version Queen's University Belfast - Research Portal: Link to publication record in Queen's University Belfast Research Portal Publisher rights Copyright 2019 IEEE. This work is made available online in accordance with the publisher’s policies. Please refer to any applicable terms of use of the publisher. General rights Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the Research Portal that you believe breaches copyright or violates any law, please contact [email protected]. Download date:07. Oct. 2021 Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco-system Anh Vu Vo Chamin Nalinda Lokugam Hewage Gianmarco Russo School of Computer Science School of Computer Science Department of Computer Science University College Dublin University College Dublin University of Salerno Dublin, Ireland Dublin, Ireland Fisciano, Italy [email protected] [email protected] [email protected] Neel Chauhan Debra F.
    [Show full text]
  • ECE 3574: Applied Software Design
    ECE 3574: Applied Software Design Message Serialization Today we are going to see various techniques for serializing objects, converting them to/from byte streams. The ability to exchange objects over files, pipes, sockets, and shared memory is a common task in concurrent programming. I Serialization (Marshalling or Pickling) I Deserialization (Unmarshalling or UnPickling) I QDataStream I Google Protocol buffers In general how do we make an object in memory persistent? We write it to a file. A file is a linear stream of bytes that may be I text-based: human-readable/editable, very portable and future proof, but we have to write parsers, they can be inefficient, can loose precision, take care to handle utf-8 I binary: machine-readable only, have to take care to make portable (Endianess), can easily get lost over time, but can preserve precision To recover the object we read the file back in. Example: ASCII text-based file Suppose we want to store some information about a task in some Todo app. struct Task { std::string description; bool completed; char priority; std::string context; }; Let’s store this as an ASCII text file. See example code: task_text_delimited.cpp There are a few standard text file formats I delimited files: common delimiters ‘,’, ‘:’, ‘|’ I INI files (originally from windows) I XML (heavy, usually overkill) I JSON (similar to xml but lighter) JSON can handle UTF encoding, arrays, many data types, and there are light-weight open-source parsing libraries for it (I like jsoncpp). Example: binary formatted file Binary files use sequences of bytes to store the same representation as in memory.
    [Show full text]
  • HDF5 Portable Parallel - Juelich.De I/O and I/O Data Formats Member of the Helmholtz-Association March March 14Th, 2017 Outline
    Parallel I/O and Portable Data Formats HDF5 Sebastian Lührs [email protected] Association - Jülich Supercomputing Centre Forschungszentrum Jülich GmbH Jülich, March 14th, 2017 Member Member of the Helmholtz Outline . Introduction . Structure of the HDF5 library . Terms and definitions . HDF5 - programming model and API . Creating/opening HDF5 files . Closing HDF5 files and other objects . HDF5 predefined datatypes . Creating dataspaces . Creating datasets . Writing/reading data . Row major / column major Association - . Partial I/O . Parallel HDF5 Member Member of the Helmholtz March 14th, 2017 2 What is HDF5? Hierarchical Data Format . API, data model and file format for I/O management . Tools suite for accessing data in HDF5 format Association - Member Member of the Helmholtz March 14th, 2017 3 HDF5 - Features . Supports parallel I/O . Self describing data model which allows the management of complex data sets . Portable file format . Available on a variety of platforms . Supports C, C++, Fortran 90 and Java . Pythonic interfaces also available . Provides tools to operate on HDF5 files and data Association - Member Member of the Helmholtz March 14th, 2017 4 Layers of the HDF5 Library Application Your Application Object API (C, F90, C++, Java, Python) Library Internals split HDF5 Library stdio MPI I/O Custom files Virtual File I/O Association - Storage device Member Member of the Helmholtz March 14th, 2017 5 File organization . HDF5 file structure corresponds in many respects to a Unix/Linux file system (fs) HDF5 Unix/Linux fs Group ↔ Directory Data set ↔ File Association - Member Member of the Helmholtz March 14th, 2017 6 Terminology File Container for storing data Group Structure which may contain HDF5 objects, e.g.
    [Show full text]